Table 1.
Summary of studies relating hyperglycaemia on admission to hospital with outcome in various conditions
| Medical condition | Country & Reference | Study years | Type of study | Patients (n) | Male: Female ratio (%) | Mean age (years) (SD) | Setting | Definition of hyperglycaemia (mmol/L) | Outcome |
|---|---|---|---|---|---|---|---|---|---|
| Myocardial infarction | Canada3 | 1966-1998 | Meta-analysis | 3387 | - | - | Hospital MI patients | Variable* | 3.9 times mortality risk with hyperglycaemia |
| Sweden4 | 1995-1997 | Retrospective with prospective follow-up | 197 | 72:28 | 68±12 | Consecutive MI admissions to ICU | Mean plasma glucose 8.1±3.0 | 30% mortality at first hospitalization and 15% during the follow up period | |
| France15 | 2000-2001 | Prospective | 146 | 68:32 | 63±15 | Consecutive MI admissions | Not-defined | Admission glucose level higher in those who died (11.7 vs 8.0 mmol/L, P=0.002) | |
| Holland13 | 1989-1996 | Retrospective and prospective | 846 | 70:30 | 65±12 | Consecutive MI admissions to CCU | Plasma glucose > 11.1 | Increase of 1 mmol/L plasma glucose associated with 4% mortality increase (non-diabetic) and 5% mortality increase (diabetic) | |
| Estonia12 | 2001-2002 | Retrospective | 779 | 59:41 | 68±12 | Consecutive MI admissions | Plasma glucose > 11.0 | 180-day mortality was 48% in patients with plasma glucose > 11.0 vs 14% in those with plasma glucose = 11.0 (P<0.0001) | |
| Germany14 | 1991-1997 | Prospective | 314 | - | - | Consecutive MI admissions | Not defined | Mortality risk increase of 1.42 in diabetic and 1.54 in non-diabetic patients for each 2.6 mmol/L rise in admission plasma glucose. | |
| Stroke | UK2 | 1990-1993 | Prospective with long-term follow-up | 811 | 50:50 | 70 (Median) | Stroke admissions with outpatient follow-up | Plasma glucose > 8.0 | 3 months mortality: 22% overall, 40% in patients with hyperglycaemia (P=0.0003) |
| Denmark20 | 1991-1993 | Prospective | 1169 | 47:53 | 74±11 | Acute stroke admissions | Plasma glucose > 11.0 | 17% mortality in non-diabetic patients, 24% in known diabetic patients, 32% in patients with new diabetes (P=0.03) | |
| Italy18 | 1984 | Prospective with 30 days follow-up | 72 | 68:32 | 68 (38-91) | Consecutive stroke admissions | Fasting serum glucose > 6.1 | 78% mortality in non-diabetic patients, 45% in diabetic patients, 29% in normoglycaemic patients (P<0.001) | |
| Canada22 | 1966-2000 | Meta-analysis | 3068 | - | - | Hospitalized patients with stroke | Variable* | Patients with hyperglycaemia and ischaemic (but not haemorrhagic) stroke had increased mortality risk (RR 3.3; 95% CI 2.3-4.8) | |
| Other diseases | USA1 | 1998 | Retrospective | 2030 | 38:62 | 59±5 | General hospital admissions | Fasting blood glucose≥7.0 Random blood glucose≥11.1 | 16% mortality in unknown diabetic patients, 3% in known diabetic patients, 1.7% in normoglycaemic patients (P<0.01) |
| Canada26 | 2000-2002 | Prospective (multicentre) | 2471 | 48:52 | 75 | Consecutive pneumonia admissions | Plasma glucose > 11.0 | 13% mortality with blood glucose > 11.0 mmol/L vs 9% with blood glucose ≤ 11.0 mmol/L (P=0.03) |
CCU: Coronary Care Unit; ICU: Intensive Care Unit; MI: Myocardial Infarction
*
Hyperglycaemia definitions varied between studies analysed